Autonomous Tracked Robots in Planar Off-Road Conditions: Modelling, Localization, and Motion Control

122 Pages·2014·6.717 MB·English

by  Ramón González, Francisco Rodríguez, José Luis Guzmán (auth.)

Studies in Systems, Decision and Control 6
Ramón González
Francisco Rodríguez
José Luis Guzmán
Autonomous Tracked
Robots in Planar
Off -Road Conditions
Modelling, Localization, and Motion Control
Studies in Systems, Decision and Control
Volume 6
Serieseditor
JanuszKacprzyk,PolishAcademyofSciences,Warsaw,Poland
e-mail:[email protected]
Forfurthervolumes:
http://www.springer.com/series/13304
AboutthisSeries
The series "Studies in Systems, Decision and Control" (SSDC) covers both new
developmentsand advances, as well as the state of the art, in the variousareas of
broadly perceived systems, decision making and control- quickly, up to date and
with a high quality. The intent is to cover the theory, applications, and perspec-
tives on the state of the art and future developmentsrelevant to systems, decision
making,control,complexprocessesandrelatedareas,asembeddedinthefieldsof
engineering,computerscience,physics,economics,socialandlifesciences,aswell
astheparadigmsandmethodologiesbehindthem.Theseriescontainsmonographs,
textbooks,lecturenotesandeditedvolumesinsystems, decisionmakingandcon-
trol spanning the areas of Cyber-PhysicalSystems, AutonomousSystems, Sensor
Networks,ControlSystems,EnergySystems,AutomotiveSystems,BiologicalSys-
tems,VehicularNetworkingandConnectedVehicles,AerospaceSystems,Automa-
tion, Manufacturing, Smart Grids, Nonlinear Systems, Power Systems, Robotics,
SocialSystems,EconomicSystemsandother.Ofparticularvaluetoboththecon-
tributorsandthereadershiparetheshortpublicationtimeframeandtheworld-wide
distribution and exposure which enable both a wide and rapid dissemination of
researchoutput.
·
Ramón González Francisco Rodríguez
José Luis Guzmán
Autonomous Tracked
Robots in Planar
Off-Road Conditions
Modelling, Localization, and Motion Control
ABC
RamónGonzález JoséLuisGuzmán
DepartmentofComputerScience DepartmentofComputerScience
UniversityofAlmeria UniversityofAlmeria
Ctra.Sacramentos/n Ctra.Sacramentos/n
04120Almeria 04120Almeria
Spain Spain
E-mail:[email protected]
FranciscoRodríguez
DepartmentofComputerScience
UniversityofAlmeria
Ctra.Sacramentos/n
04120Almeria
Spain
Additionalmaterialtothisbookcanbedownloadedfromhttp://extras.springer.com
ISSN2198-4182 ISSN2198-4190 (electronic)
ISBN978-3-319-06037-8 ISBN978-3-319-06038-5 (eBook)
DOI10.1007/978-3-319-06038-5
SpringerChamHeidelbergNewYorkDordrechtLondon
LibraryofCongressControlNumber:2014935086
(cid:2)c SpringerInternationalPublishingSwitzerland2014
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof
thematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,
broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformation
storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodology
nowknownorhereafterdeveloped.Exemptedfromthislegalreservationarebriefexcerptsinconnection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered
and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of
this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’slocation,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.
PermissionsforusemaybeobtainedthroughRightsLinkattheCopyrightClearanceCenter.Violations
areliabletoprosecutionundertherespectiveCopyrightLaw.
Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication
doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant
protectivelawsandregulationsandthereforefreeforgeneraluse.
Whiletheadviceandinformationinthisbookarebelievedtobetrueandaccurateatthedateofpub-
lication,neithertheauthorsnortheeditorsnorthepublishercanacceptanylegalresponsibilityforany
errorsoromissionsthatmaybemade.Thepublishermakesnowarranty,expressorimplied,withrespect
tothematerialcontainedherein.
Printedonacid-freepaper
SpringerispartofSpringerScience+BusinessMedia(www.springer.com)
You have to test your stuff a lot,
test it just like you’re going to fly it,
and then fly it just like you tested it.
Professor Steven W. Squyres
Principal Investigator
Mars Exploration Rover mission
Contents
1 Introduction ......................................... 1
1.1 The Importance of Tracked Vehicles ..................... 1
1.1.1 Tracked Vehicles along History.................... 1
1.1.2 Autonomous Tracked Vehicles..................... 2
1.2 Motivation and Contributions........................... 5
1.3 Outline of this Monograph.............................. 8
1.4 Assumptions and Limitations ........................... 8
2 Modelling Tracked Robots in Planar Off-Road
Conditions........................................... 11
2.1 Introduction .......................................... 11
2.2 Extended Kinematic Model with Slip .................... 14
2.3 Extended Trajectory Tracking Error Model with Slip....... 16
2.4 Discrete-Time Trajectory Tracking Error Model and
Model Uncertainties ................................... 20
2.5 State and Input Constraints ............................ 21
2.6 Estimating Slip ....................................... 22
2.7 Results............................................... 23
2.7.1 Testing the Sensor Performance ................... 23
2.7.2 Model Validation................................ 25
2.7.3 Additive Uncertainty Identification ................ 29
2.8 Conclusions........................................... 32
3 Localization of Tracked Robots in Planar Off-Road
Conditions........................................... 35
3.1 Introduction .......................................... 35
3.2 Localization Using Visual Odometry ..................... 37
3.2.1 Template Matching.............................. 39
3.2.2 Estimating Robot Displacement................... 41
3.2.3 Estimating Robot Orientation: Visual Compass ..... 42
3.2.4 LocalizationApproachCombiningVisualOdometry
with Visual Compass ............................ 44
3.2.5 Computational Aspects of Template Matching ...... 44
VIII Contents
3.3 Results............................................... 47
3.3.1 Testing the Sensor Performance ................... 47
3.3.2 Localization Strategies Validation ................. 49
3.4 Conclusions........................................... 56
4 Adaptive Motion Controllers for Tracked Robots ....... 57
4.1 Introduction .......................................... 57
4.2 Slip Compensation Adaptive Controller Using Dynamic
Feedback Linearization................................. 59
4.3 Slip Compensation Adaptive Controller Using an
LMI-Based Approach .................................. 61
4.3.1 Problem Statement .............................. 63
4.3.2 Asymptotic Stability and Performance ............. 65
4.3.3 Input Constraints ............................... 65
4.3.4 State Constraints................................ 67
4.3.5 Performance Region ............................. 68
4.3.6 Final Optimization Problem ...................... 68
4.4 Results............................................... 72
4.4.1 Experimental Results ............................ 73
4.5 Conclusions........................................... 82
5 Robust Predictive Motion Controller for Tracked
Robots .............................................. 85
5.1 Introduction .......................................... 85
5.2 Problem Statement .................................... 87
5.3 Robust Tube-Based MPC Controller ..................... 88
5.3.1 Local Compensation of System Dynamics .......... 90
5.3.2 Reachable Sets Calculation ....................... 91
5.3.3 Terminal Constraints for MPC .................... 92
5.3.4 MPC Strategy .................................. 93
5.4 Results............................................... 94
5.4.1 Experiment 1. Circular Trajectory................. 94
5.4.2 Experiment 2. U-shaped Trajectory................ 99
5.5 Conclusions........................................... 104
6 Conclusions and Future Works ........................ 107
6.1 Conclusions........................................... 107
6.2 Future Works ......................................... 109
References............................................... 111
Chapter 1
Introduction
This chapter addresses a historical perspective of tracked vehicles
and motivates the use of tracked mobile robots. It highlights the
main contributions of this monograph. Finally, the main assump-
tions and limitations regarding the developed work are explained.
1.1 The Importance of Tracked Vehicles
1.1.1 Tracked Vehicles along History
Theoriginsoftrackedvehiclesarecloselymotivatedbytheideaofdeveloping
cross-country vehicles, that is, vehicles moving on off-road rough terrains
(packed snow, muddy roads, loose sandy soils, etc.). In 1770, a patent by
RichardL. Edgeworthfirstly introducedthe conceptofwhat is knowntoday
as full-track [6]: “the invention cossets in making portable railways to wheel
carriages so that several pieces of wood are connected to the carriage which
it moves in regular succession in such a manner that a sufficient length of
railing is constantly at rest for the wheels, to roll upon, and that when the
wheelshavenearlyapproachedtheextremityofthispartoftherailway,their
motionsshalllaydownafreshlengthofrailinfront,theweightofwhichinits
descent shall assist in raising such part of the rail as the weeks have already
passed over, and thus the pieces of wood which are taken up in the rear are
in succession laid in the front, so as to furnish constantly a railway for the
weeks to roll on.” By those years none of the inventions became successful
products. A further step came with the introduction of internal-combustion
engines at the end of the 19th century. In this sense, a clear contribution
was the Holt tractor in 1911 (Figure 1.1a). This vehicle was widely used by
the British, French, and American armies in World War I for hauling heavy
equipment. This successful platform led to the development of war tanks.
After the War, the prolific research carried out was applied to agricul-
ture. Indeed during the years 1920-1940, many fundamental questions were
unfolded such as: the relationship between track and soil, the movement re-
R.Gonza´lez,F.Rodr´ıguez,andJ.L.Guzma´n,Autonomous TrackedRobotsin Planar 1
Off-RoadConditions,StudiesinSystems,DecisionandControl6,
DOI:10.1007/978-3-319-06038-5_1, (cid:2)c SpringerInternationalPublishingSwitzerland2014
2 1 Introduction
sistance, sinkage, wheel dimensions, “ground pressure”, etc. That research
led to consider the development of tracked vehicles as a solid discipline [6].
Atthispoint,itisinterestingtointroduceabriefdiscussioncomparingthe
performance of wheels and tracks. In this regardthe article [133] constitutes
asolidreference.There,afterseveralcomparisonsondifferentterrains(sand,
clay, loam), authorsdemonstrate that the thrust developedby a wheeled ve-
hicle will generally be lower than that developed by a comparable tracked
vehicle, this is particularly noticeable on cohesive soils. This is motivated by
the fact that the average normal pressure under the tires of a wheeled vehi-
cle will be higher than that under a comparable tracked vehicle. The paper
[54] states that the wheeled-vs-track dilemma is rooted in the following fun-
damental subjects: the vehicle’s mission, the terrain profile, and the specific
vehicularfeatures.Inparticular,themainadvantagesoftrackedvehiclesare:
• They offer the best solution for a versatile platform that is required to
operateoverdiverseterrainsevenatdifferentweatherconditions.Forthat
reason this kind of vehicles are very common in search and rescue opera-
tions especially on snow.
• They generate low ground pressure, which leads to conserve the natural
environment.Thismotivateswhytrackedvehiclesareemployedinagricul-
ture(harvesting,planting,orsprayingtasks)andminingactivities(Figure
1.1b, d).
• They prevent from sinking, even becoming stuck, into soft ground, there-
fore they are ideal vehicles for loose sandy terrains like in military opera-
tions(haulingheavyartilleryoraspersonneltransportationunits)(Figure
1.1a, c).
1.1.2 Autonomous Tracked Vehicles
It is clear that trackedvehicles comprise a solid and successful way of trans-
portation,especiallyinoff-roadconditions.Forthatreason,whenaroboticist
thinks about a mobile robotto operate in off-roadconditions one of the first
optionsistousetrackedlocomotion.Howeverthetaskofmakingautonomous
a (tracked) vehicle leads to consider many challenging issues.
Thebehaviourofamobilerobotinoutdooroff-roadconditionsisquitedi-
fferentfrom a mobile robotworkingonstructured indoorenvironments. The
most important disturbances and inconveniences affecting to off-road mo-
bile robots deal with the robot-terrain interactions, such as slip and sinkage
phenomena.Thefirstissuetobesolvedisrelatedtotherobotmechanics.Off-
road applications generally require robots to travel across unprepared rough
terrains. This fact means that proper locomotion mechanisms and on-board
sensors should be selected to safely steer the robot autonomously [53, 117].
Furthermore, due to the robot can work in remote or inaccessible scenarios,
efficient power units and reliable communication systems are also required.
Figure1.2showsseveraltrackedmobilerobotsappliedinchallengingoff-road